Add option to write fragment level quant #443
Conversation
alphadia/constants/default.yaml
Outdated
| @@ -237,6 +237,7 @@ fdr: | |||
| search_output: | |||
| peptide_level_lfq: false | |||
| precursor_level_lfq: false | |||
| fragment_quant_matrix: false | |||
There was a problem hiding this comment.
can you add a comment what this is doing and that this is an advanced feature
alphadia/outputtransform.py
Outdated
| write_df( | ||
| group_intensity_df, | ||
| os.path.join( | ||
| self.output_folder, f"fragment_{level_name}filt.matrix" |
There was a problem hiding this comment.
yes, it is meant to indicate that the fragments are filtered using the precursor-level fragment filter
There was a problem hiding this comment.
please be more explicit, _filtered or smth
tests/e2e_tests/e2e_test_cases.yaml
Outdated
| @@ -85,6 +85,7 @@ test_cases: | |||
| search_output: | |||
| peptide_level_lfq: true | |||
| precursor_level_lfq: true | |||
| fragment_quant_matrix: true | |||
There was a problem hiding this comment.
Let's only test this in one of the testcases
alphadia/outputtransform.py
Outdated
| "peptide", | ||
| self.config["search_output"]["peptide_level_lfq"], | ||
| ), | ||
| quantlevel_configs = [ # the configs specified here are: quantlevel, level_name, should_process, save_fragments |
There was a problem hiding this comment.
this could be a namedtuple for instance, then the comment will be obsolete as the code will tell
There was a problem hiding this comment.
dataclass should be nicer:
@dataclass
class LFQOutputConfig:
should_process: bool
quant_level: str
level_name: str
save_fragments: bool = False
quantlevel_configs = [
LFQOutputConfig(
self.config["search_output"]["precursor_level_lfq"],
"mod_seq_charge_hash",
"precursor",
self.config["search_output"]["fragment_quant_matrix"],
),
LFQOutputConfig(
self.config["search_output"]["peptide_level_lfq"],
"mod_seq_hash",
"peptide",
),
LFQOutputConfig(
True,
"pg",
"pg",
),
]
lfq_results = {}
for quantlevel_config in quantlevel_configs:
if not quantlevel_config.should_process:
continue
....
alphadia/outputtransform.py
Outdated
| write_df( | ||
| group_intensity_df, | ||
| os.path.join( | ||
| self.output_folder, f"fragment_{level_name}filt.matrix" |
There was a problem hiding this comment.
please be more explicit, _filtered or smth
alphadia/constants/default.yaml
Outdated
| @@ -237,6 +237,7 @@ fdr: | |||
| search_output: | |||
| peptide_level_lfq: false | |||
| precursor_level_lfq: false | |||
| fragment_quant_matrix: false #advanced feature to write out quantitative matrix with fragment ion intensities e.g. for AlphaQuant | |||
There was a problem hiding this comment.
save_fragment_quant_matrix?
should this also be part of the GUI @GeorgWa ?
There was a problem hiding this comment.
Renamed it. Indeed I think it should be part of the GUI, would be good for AlphaQuant compatibility.
| @@ -120,6 +121,7 @@ test_cases: | |||
| search_output: | |||
| peptide_level_lfq: true | |||
| precursor_level_lfq: true | |||
| save_fragment_quant_matrix: false | |||
There was a problem hiding this comment.
given that false is the default, no need to specify it here in the yamls
| group_intensity_df, | ||
| os.path.join( | ||
| self.output_folder, | ||
| f"fragment_{quantlevel_config.level_name}filtered.matrix", |
There was a problem hiding this comment.
is .matrix a special file format? or could this be something like .matrix.tsv?
There was a problem hiding this comment.
I think the file format is set by write_df. To me it would also be fine if the advanced output is always saved as .parquet
There was a problem hiding this comment.
ah, did not look into write_df to see path_no_format 👍
| group_intensity_df, | ||
| os.path.join( | ||
| self.output_folder, | ||
| f"fragment_{quantlevel_config.level_name}filtered.matrix", |
There was a problem hiding this comment.
I think the file format is set by write_df. To me it would also be fine if the advanced output is always saved as .parquet
This PR adds the option to write to disk a matrix with fragment-level quantities which can be used for detailed inspection of the data or downstream quantification algorithms.
Changes:
SearchPlanOutput.build_lfq_tables()to support fragment-level table export. When enabled, the feature exports the precursor-levelgroup_intensity_dfcontaining all fragments that pass the precursor level correlation filtersconfig_dict["search_output"]["fragment_quant_matrix"](default: false)